Install on Kubernetes with Helm

This page explains how to install Kong Gateway with Kubernetes Ingress Controller using Helm.

  • The Enterprise deployment includes a Postgres sub-chart provided by Bitnami.
  • For open-source deployments, you can choose to use the Postgres sub-chart, or install without a database.

Configuration for both options is flexible and depends on your environment.

The documentation on installing with a flat Kubernetes manifest also explains how to install in DB-less mode for both Enterprise and OSS deployments.

The Kong Gateway software is governed by the Kong Software License Agreement. Kong Gateway (OSS) is licensed under an Apache 2.0 license.

Prerequisites

  • A Kubernetes cluster, v1.19 or later
  • kubectl v1.19 or later
  • (Enterprise only) A license.json file from Kong
  • Helm 3

Create namespace

Create the namespace for Kong Gateway with Kubernetes Ingress Controller. For example:

  1. kubectl create namespace kong

Set up Helm

  1. Add the Kong charts repository:

    1. helm repo add kong https://charts.konghq.com

  2. Update Helm:

    1. helm repo update

Create license secret

  1. Save your license file temporarily with the filename license (no file extension).

  2. Run:

    1. kubectl create secret generic kong-enterprise-license --from-file=./license -n kong

Create secret for RBAC superuser (recommended)

If you plan to use RBAC, you must create a secret for the superuser account password at this step in installation. You cannot create it later.

  1. Create the RBAC account.

  2. Create the secret:

    1. kubectl create secret generic kong-enterprise-superuser-password \
    2. -n kong \
    3. --from-literal=password={YOUR_PASSWORD}

Create secret for Session plugin

If you create an RBAC superuser and plan to work with Kong Manager or Dev Portal, you must also configure the Session plugin and store its config in a Kubernetes secret:

  1. Create a session config file for Kong Manager:

    1. $ echo '{"cookie_name":"admin_session","cookie_samesite":"off","secret":"<your-password>","cookie_secure":false,"storage":"kong"}' > admin_gui_session_conf

  2. Create a session config file for Kong Dev Portal:

    1. $ echo '{"cookie_name":"portal_session","cookie_samesite":"off","secret":"<your-password>","cookie_secure":false,"storage":"kong"}' > portal_session_conf

    Or, if you have different subdomains for the portal_api_url and portal_gui_host, set the cookie_domain and cookie_samesite properties as follows:

    1. $ echo '{"cookie_name":"portal_session","cookie_samesite":"off","cookie_domain":"<.your_subdomain.com">,"secret":"<your-password>","cookie_secure":false,"storage":"kong"}' > portal_session_conf

  3. Create the secret:

    For Kong Manager only:

    1. kubectl create secret generic kong-session-config \
    2. -n kong \
    3. --from-file=admin_gui_session_conf

    For Kong Manager and Dev Portal:

    1. kubectl create secret generic kong-session-config \
    2. -n kong \
    3. --from-file=admin_gui_session_conf \
    4. --from-file=portal_session_conf

Create values.yaml file

Create a values.yaml file to provide required values such as password secrets or optional email addresses for notifications. You can work from the Enterprise example file. The example file includes comments to explain which values you must set.

For OSS deployments, the default install might be sufficient, but you can explore other values.yaml files and the readme in the charts repository, which includes an exhaustive list of all possible configuration properties.

Note that the Enterprise deployment includes a Postgres sub-chart provided by Bitnami. You might need to delete the PersistentVolume objects for Postgres in your Kubernetes cluster to connect to the database after install.

Deploy Kong Gateway with Kubernetes Ingress Controller

  1. Run:

    1. ## Kong Gateway
    2. helm install my-kong kong/kong -n kong --values ./values.yaml
    1. ## Kong Gateway (OSS)
    2. helm install kong/kong --generate-name --set ingressController.installCRDs=false

    For more information on working with Helm charts for Kong Gateway (OSS), see the chart documentation.

    This might take some time.

  2. Check pod status, and make sure the my-kong-kong-{ID} pod is running:

    1. kubectl get pods -n kong

Finalize configuration and verify installation

  1. Run:

    1. kubectl get svc my-kong-kong-admin \
    2. -n kong \
    3. --output=jsonpath='{.status.loadBalancer.ingress[0].ip}'

  2. Copy the IP address from the output, then add the following to the .env section of your values.yaml file:

    1. admin_api_uri: {YOUR-DNS-OR-IP}

    Note: If you configure RBAC, you must specify a DNS hostname instead of an IP address.

  3. Clean up:

    1. kubectl delete jobs -n kong --all

  4. Update with changed values.yaml:

    1. helm upgrade my-kong kong/kong -n kong --values ./values.yaml

  5. After the upgrade finishes, run:

    1. kubectl get svc -n kong

    With an Enterprise deployment, the output includes EXTERNAL-IP values for Kong Manager and Dev Portal. For example:

    1. NAME                          TYPE           CLUSTER-IP       EXTERNAL-IP     PORT(S)                            AGE
    2. my-kong-kong-manager          LoadBalancer   10.96.61.116     10.96.61.116    8002:31308/TCP,8445:32420/TCP      24m
    3. my-kong-kong-portal           LoadBalancer   10.101.251.123   10.101.251.123  8003:31609/TCP,8446:32002/TCP      24m

Next steps

See the Kong Ingress Controller docs for how-to guides, reference guides, and more.